This 8-pound copper cylinder can cool an Intel Core i9 CPU pretty effectively

[midian182]

WTF?! There are plenty of unconventional ways to keep your PC's processor cool, like using a massive, 8-pound piece of copper that sticks up from the motherboard like a lighthouse, for example. It might sound ridiculous, and the method seems impractical, but it does work pretty effectively.

Redditor That-Desktop-User posted their DIY solution on the HardwareGore subreddit. The copper cylinder is being used to cool an unnamed Intel Core i9 CPU in a fully open-air system.

It sounds as if the thermally conductive material does a good job of passively cooling Intel's CPU. The i9 chips are known for getting very toasty, but the copper keeps it at 35 degrees centigrade (95 degree Fahrenheit) while idle. Running a stress test for under a minute caused the temperatures to jump to 80 Centigrade (176° Fahrenheit), and it took just ten seconds after the test was finished for the CPU to return to its idling temperature.

That-Desktop-User explains that they got the block from an old medical machine used to stimulate a person's vitals they bought for a few thousand dollars. The copper cylinder may have previously functioned as a cooler in the device as it has holes in the sides threaded for liquid flow. They believe it was originally used for nitrogen and oxygen and that it would be pretty easy to switch out the original 2-4mm (inner diameter) fittings for larger-diameter piping.

Some pieces of information are missing from the post, such as how the cylinder was fastened to the motherboard and what thermal paste was used.

Copper is a component found in many modern heatsinks, though not 8 pounds of it inside one cooler. The closest comparable product that comes to mind might be Thermalright's chunky, 4.1-pound True Copper passive heatsink from a few years ago.

Copper is a fairly expensive material, though work is being done to make it atom-thin, which could eventually be used to replace gold in chips, thereby lowering their build price.

Permalink to story.

https://www.techspot.com/news/97505-8-pound-copper-cylinder-can-cool-cpu-pretty.html

 

[ShadowDeath]

Imagine having a few thousand dollars to spend on a copper cylinder that you know little to nothing about. Lol

 

[m4a4]

Lol, I wonder if this thing can even passively cool down enough over a long, heavy workload.

This thing looks great for short bursts, but once it's thermally saturated, I suspect it won't be doing much (without actively using that hole to cool it down).

 

[oasked]

Lol, I wonder if this thing can even passively cool down enough over a long, heavy workload.

This thing looks great for short bursts, but once it's thermally saturated, I suspect it won't be doing much (without actively using that hole to cool it down).

Exactly - once the thermal mass is 'full', it won't cool at all.

 

[mbrowne5061]

Imagine having a few thousand dollars to spend on a copper cylinder that you know little to nothing about. Lol

Sounds like it was more them spending a few thousand on some obsolete medical devices, and this cylinder was a component in one of them.

That much copper is only worth about $30

 

[Avro Arrow]

I'd be concerned about the weight causing mechanical damage to the motherboard.

 

[RudyBob]

Cast iron is less expensive

 

[neeyik]

Cast iron is less expensive

Its relatively low thermal conductivity makes it a poor choice for a heatsink.

 

[p51d007]

Spot market shows aroung $4.04/pound a couple days ago.

 

[MakeMSGreatAgain]

Sounds like it was more them spending a few thousand on some obsolete medical devices, and this cylinder was a component in one of them.

That much copper is only worth about $30

5 US pound ingot of copper is going for about $60 on ebay... so maybe $100 or so worth of copper right now, possibly.

I think a tall 'U' shaped internal water cooling loop in the cylinder with hoses attached and a simple water pump to circulate water through without a fan would probably work pretty darn well. If anything it is really interesting to look at, even though it likely isn't effective for any long sustained loads.

 

[captaincranky]

Where exactly would one find a case capable of containing this monstrosity?

Besides, it will never sell, no RGB.

 

[RudyBob]

Its relatively low thermal conductivity makes it a poor choice for a heatsink.

I don't know what thermal conductivity is. My post was based on cast iron's ability to get hot and stay hot which sounds just like what the folk were saying about copper. My definition of conductivity is ability to transfer. And thermal(heat) is what is being transfered. Fill me in I would like to know. Thanks. Don't get to technical on me now

 

[neeyik]

I don't know what thermal conductivity is. My post was based on cast iron's ability to get hot and stay hot which sounds just like what the folk were saying about copper. My definition of conductivity is ability to transfer. And thermal(heat) is what is being transfered. Fill me in I would like to know. Thanks. Don't get to technical on me now

Thermal conductivity is essentially how much heat energy can travel through a fixed size of material in one second. The higher the conductivity, the quicker the all of the material will gain heat energy.

Copper’s thermal conductivity is around 5 to 10 higher than cast iron’s. For effective cooling with metal heatsinks, you want them to rise in temperature as quickly as possible, because the hotter something is, compared to its surroundings, the higher the rate of heat energy loss will be.

In other words, you want the heatsink to absorb heat energy quickly, rise in temperature quickly, then lose that heat energy to the air quickly. Copper does this much better than cast iron, as does aluminum, which is why they’re the most commonly used metals for semiconductor heatsinks.

 

[mbrowne5061]

5 US pound ingot of copper is going for about $60 on ebay... so maybe $100 or so worth of copper right now, possibly.

I think a tall 'U' shaped internal water cooling loop in the cylinder with hoses attached and a simple water pump to circulate water through without a fan would probably work pretty darn well. If anything it is really interesting to look at, even though it likely isn't effective for any long sustained loads.

That is an ingot, good to go into a manufacturing cycle, not scrap. This piece is effectively scrap, unless you need this exact part, because it'll need to be melted down, scrape off the slag/impurities, and then cast into an ingot - then it will be worth more. Right now, copper has a scrap value just south of $4/lb, so this 8lb hunk of copper is only worth around $30

 

[mbrowne5061]

I don't know what thermal conductivity is. My post was based on cast iron's ability to get hot and stay hot which sounds just like what the folk were saying about copper. My definition of conductivity is ability to transfer. And thermal(heat) is what is being transfered. Fill me in I would like to know. Thanks. Don't get to technical on me now

Cast iron is a poor conductor of heat. Your point about it "staying hot" is exactly why it makes for a poor heatsink. The goal of a heat sink is to take heat from a smaller, concentrated area, and quickly conduct it away from there so that it has access to a larger surface area.

Aluminum, Copper, and Gold are pretty much the champs when it comes to conductivity - both thermal and electrical - and which one is best really comes down to what is most important in the heat sink. Copper takes the crown for being the most thermally conductive metal (that isn't some kind of super alloy that costs an arm and a leg), but it is expensive (~$4/lb scrap, $8-10/lb 'new'). Aluminum offer still pretty good thermal conductivity, and its dirt cheap ($0.50/lb scrap, $2-4/lb 'new'). Gold is neither better than copper when it comes to conductivity, nor cheaper (duh), but it does not corrode or tarnish, still offers very good conductivity, and can be easily plated onto other metals, so you'll often see it get used as a 'top coat' on a heat sink, if corrosion is a concern.

For some context, the conductivity of cast iron is about 1/4 that of aluminum, and aluminum's conductivity is around 1/2 of copper. Gold is only slightly worse than copper in terms of conductivity. Silver is nearly as good as copper on thermal conductivity, but it is expensive and it can corrode/tarnish.

 

[Avro Arrow]

Where exactly would one find a case capable of containing this monstrosity?

Besides, it will never sell, no RGB.

Ah, just wrap it up with this:

 

[grumblguts]

Yeah until it cant

 

[captaincranky]

I don't know what thermal conductivity is. My post was based on cast iron's ability to get hot and stay hot which sounds just like what the folk were saying about copper. My definition of conductivity is ability to transfer. And thermal(heat) is what is being transfered. Fill me in I would like to know. Thanks. Don't get to technical on me now

The easiest way to explain, "thermal conductivity", is to think of it as a "wick", just not for fluids, but for heat. Which is why copper and aluminum are used since their ability to draw, "or wick", heat away from a source is the best of all common metals.

Copper is heavy and expensive, but the best at "wicking" heat away from a source. which is why the better modern air coolers use the metal in direct contact with the CPU, then into aluminum fins..

Aluminum doesn't cool as well as copper but, the surface area of the cooler fins makes up for that deficiency. Were you to calculate the surface area of one fin, multiply it by the number of fins in the cooler, and then multiply that by 2 (??), you would realize how much aluminum is exposed to the air flow.

As for cast iron, it's incredibly brittle, thus unsuitable for casting into objects as thin as cooler fins. Its mediocre ability to cool, is the reason automobile drum brakes "fade", and subsequently caused disc brakes to be installed on all cars after ????. (You'll have to look that up, but I'm sure ti was 1975 or before).

Disc brakes respond in a linear relationship with pedal pressure, and the braking surface is exposed directly to the air. Hence, you get more predictable response and better cooling.

Anecdotally, in the Pentium, "Prescott" era, stock Intel coolers had copper cores. Had we have saved them, they would have made decent coolers for many of today's low power CPUs. They certainly would have been better than the trash Intel was/is giving away in recent times

 

[mbrowne5061]

As for cast iron, it's incredibly brittle, thus unsuitable for casting into objects as thin as cooler fins. Its mediocre ability to cool, is the reason automobile drum brakes "fade", and subsequently caused disc brakes to be installed on all cars after ????. (You'll have to look that up, but I'm sure ti was 1975 or before).

Depends on the model. I know Toyota 4runners had rear drum brakes up until at least 2002 (and discs in the front). Doing a "Taco upgrade" was a popular mod; Tacomas of the same generation used Disc brakes in the front & rear, the fronts were larger than what were installed in 4runners, and the OEM parts for both front & rears fit as-is in the 4runner.

Pretty sure the 5th generation 4runners use disc brakes everywhere, though.

 

[captaincranky]

Depends on the model. I know Toyota 4runners had rear drum brakes up until at least 2002 (and discs in the front). Doing a "Taco upgrade" was a popular mod; Tacomas of the same generation used Disc brakes in the front & rear, the fronts were larger than what were installed in 4runners, and the OEM parts for both front & rears fit as-is in the 4runner.

I was a bit vague on my description of , "when disc brakes were first installed". I should have said, "front disc brakes". My 1975 Pontiac Le Mans was equipped with front discs, and rear drums. This configuration was pretty much standard from that point onward.. Four wheel discs, (AFAIK), were mostly available on Corvettes at around the same time.

The Toyota phenomenon you describe seems rare, although I can't be certain that's a blanket assessment.

My 1997 Suzuki "Sidekick", is configured with front discs and rear drums. So, the Four Runner seems like it might be the "odd man out". (Again, I'm guessing, since I haven't worked in that industry for decades).

 

[mbrowne5061]

I was a bit vague on my description of , "when disc brakes were first installed". I should have said, "front disc brakes". My 1975 Pontiac Le Mans was equipped with front discs, and rear drums. This configuration was pretty much standard from that point onward.. Four wheel discs, (AFAIK), were mostly available on Corvettes at around the same time.

The Toyota phenomenon you describe seems rare, although I can't be certain that's a blanket assessment.

My 1997 Suzuki "Sidekick", is configured with front discs and rear drums. So, the Four Runner seems like it might be the "odd man out". (Again, I'm guessing, since I haven't worked in that industry for decades).

I think it was more just Toyota being Toyota. They fix their broken things, but don't fix something if it isn't broken. Especially not the 4runner line. They seem perfectly happy to let those things keep selling themselves, and not provide any updates (especially the 5th gens, which is about a decade old now, with near zero changes)
For the 3rd gen 4runner, it started production in 1995, and I'm around 65% sure the brakes for the third gen were copy+pasted from the second gen.

 

[captaincranky]

I think it was more just Toyota being Toyota. They fix their broken things, but don't fix something if it isn't broken. Especially not the 4runner line. They seem perfectly happy to let those things keep selling themselves, and not provide any updates (especially the 5th gens, which is about a decade old now, with near zero changes)

I misread your post, and was thinking the 4 Runner had drum brakes all around. Disc front and drum rear is the standard configuration, industry wide. (??) So, as the young hipsters say, "my bad".

FWIW, older drum "Bendix" style brakes were self activating, yielding a more logarithmic, as opposed to linear response, to pedal pressure. Which is why, "in the good old days", we were slipping, sliding, screeching, and leaving skid marks, yet still not stopping Enter "ABS" (anti-lock brakes) in later year's drum brakes. My Sidekick has this feature, as I'm reasonably sure, that Four Runners are likely so equipped as well..I think anti lock- (drum), brakes are federally mandated at present.. .

 

[ChipBoundary]

I'd be concerned about the weight causing mechanical damage to the motherboard.

Given that it takes roughly 30lbs of pressure to slot a memory DIMM and some CPU coolers apply even more, I highly doubt it.

 

[captaincranky]

Given that it takes roughly 30lbs of pressure to slot a memory DIMM and some CPU coolers apply even more, I highly doubt it.

But that 30 pounds of pressure only lasts for a split second, (or thereabouts).

A mobo will sag due to constant pressure over time, much the same as a floor joist after a few years.

I forget where I saw this, but one manufacturer's recommendation was the cooler should weigh less than one pound. The ubiquitous Cooler Master 212s already weight double that.

 

[ChipBoundary]

But that 30 pounds of pressure only lasts for a split second, (or thereabouts).

A mobo will sag due to constant pressure over time, much the same as a floor joist after a few years.

I forget where I saw this, but one manufacturer's recommendation was the cooler should weigh less than one pound. The ubiquitous Cooler Master 212s already weight double that.

You could slot those DIMM's over and over again and not see any damage to the board itself. You'd see wear on the slot or the DIMM's before the board itself. Also, the hyper 212 evo is 1.3 lbs, not what you're suggesting. I've been using one for 7 years, consistently, and my board has had zero effect. The mounting pressure of a cooler far exceeds the weight you're talking about as well.

Also, when you have things like the 4090 weighing in at 4.9 lbs hanging from the board/putting pressure on it (depending on mobo orientation/case design), I highly doubt there will be issues. Heck, a 1080ti is pushing almost 3 lbs as well.

Even if said article you mention exists, I'm gonna say with some pretty strong confidence by years of evidence to the contrary that they are quite soundly wrong.